Mac M1 optimizations, fix train pipeline, add Hey Cosmo wake word model

- Fix install_mac.sh: use venv + Python 3.12 (3.14 incompatible with ML libs)
- Fix run_mac.sh: activate venv, add CPU thread optimization env vars
- Fix agent.py: remove f-string from SYSTEM_PROMPT template (NameError on import)
- Add missing deps: sounddevice, pydub, imageio-ffmpeg, omegaconf
- Optimize for M1: torch.inference_mode, set_num_threads, OMP/MKL tuning
- Switch to qwen2.5:3b for faster LLM responses on Mac
- Switch Whisper to medium model with auto compute (small+int8 had poor Russian)
- Add initial_prompt for better Russian transcription
- Add open_app tool for native macOS app launching
- Fix TTS: sanitize Latin text to Cyrillic for Silero compatibility
- Fix wake word echo: add cooldown after TTS, reset model state, raise threshold
- Make "Слушаю" TTS synchronous to avoid mic interference
- Fix train Dockerfile: remove tensorflow/onnx2tf (only ONNX needed), fix deps
- Fix train.sh: use wget for dataset download, add --shm-size=2g
- Add trained hey_cosmo.onnx wake word model
- Add TODO section to CLAUDE.md (ChatterBox TTS, Ollama Modelfile ideas)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

config/config_mac.yaml

@@ -7,18 +7,22 @@ audio:
   silence_duration: 1.0
 
 whisper:
-  model_size: "small"        # На Mac без GPU — small быстрее чем distil-large
+  model_size: "medium"       # medium — лучшее качество русского на CPU (small слишком много ошибок)
   device: "cpu"              # Mac Intel/Apple Silicon — CPU (MPS пока не стабилен в faster-whisper)
-  compute_type: "int8"       # int8 быстрее на CPU
+  compute_type: "auto"       # auto вместо int8 — int8 слишком сильно режет качество русского
   language: "ru"
+  initial_prompt: "Cosmo, открой браузер, найди программу, запусти приложение."  # подсказка для русского контекста
 
 ollama:
   base_url: "http://localhost:11434"
-  model: "qwen2.5:7b"
+  model: "qwen2.5:3b"           # 3b быстрее на M1 (~2x), достаточно для голосовых команд
   temperature: 0.2
   max_tokens: 1024
   max_agent_steps: 10
 
+performance:
+  num_threads: 4                 # CPU потоки для torch (TTS/Whisper)
+
 tts:
   enabled: true
   silero_speaker: "eugene"   # xenia (женский) baya aidar eugene kseniya